2 50 



IsTATURE 



{Jan. 



12, I 



approximation to truth. Moreover, the method of deter- 

 mining the time of vibration of the magnet as generally 

 practised by English observers also appears to us to be 

 preferable to that adopted by M. Moureaux, although this 

 has the advantage of occupying little time and therefore 

 of minimizing the effect of any alteration in temperature 

 during these observations and those of the deflections. 



As regards inclination, there can, we think, be little 

 doubt that the Kew pattern of dip circle, as made by 

 Dover, is distinctly preferable to that used in the French 

 survey. Indeed in the latter instrument it would seem 

 to be difficult to avoid draughts and dust, the two great 

 enemies to accuracy in field work. Only one needle of 

 o'o65m. in length was used by M. Moureaux, and the 

 memoir gives no direct evidence of the degree of accuracy 

 of which it was capable. Still M. Moureaux's instrument 

 has the merit of portability, since, when packed in its box, 

 it weighs less than 2 kilos. 



As regards the plan of operations, we cannot speak too 

 highly. Every care seems to have been taken, by a pre- 

 liminary study of the ground, to select stations which 

 should be as free as possible from any local disturbance, 

 such as the proximity of railway-lines, manufactories, &c. 

 It would, however, have added to the completeness of INI. 

 Moureaux's work if to the description of the stations 

 there had been given some account of their general 

 geological character and of that of the districts in the 

 immediate neighbourhood, since, as is well known, the 

 presence of igneous rocks or of rocks containing magnetic 

 oxide of iron is the chief cause of local disturbance. 



M. Moureaux began operations at each station, as a 

 rule, at the commendably early hour of 7 a.m., so as 

 to secure the determination of the magnetic meridian 

 when the diurnal variation in declination was at about its 

 morning minimum and nearly stationary. The observa- 

 tions for the geographical meridian were made between 

 8.30 and 9 a.m. ; that is, at about the best period for the 

 observation. The determination of the horizontal com- 

 ponent was next made, a set of swings being taken before 

 and after the deflection observations, all of which were 

 completed by about 10.30 a.m. Between this time and 

 noon was occupied in the dip observations. When the 

 circumstances of travel or of weather made a departure 

 from this plan necessary, the observations of declination 

 were made either at the time of maximum of diurnal 

 variation or at about the time of evening minimum— say 

 between 5 anrl. 6 p.m. 



The results of the various observations are presented 

 with that elegance and clearness which is characteristic 

 of the publications of the Bureau Central Meteorologique. 

 They are all referred to the Pare Saint- Maur as a base 

 station, by direct comparisons with the photographic 

 curves of the registering apparatus at work in the mag- 

 netic observatory • and are reduced to the mean epoch 

 January i, 1885, by adding the difference between the 

 values obtained at the different stations and Pare Saint- 

 Maur at the time of observation to the corresponding 

 values at Pare Saint- Maur on January i, 1885, obtained 

 from the mean of the observations made there in 

 December 1884 and January 1885. This method pre- 

 supposes that the diurnal variation is of the same order 

 throughout the whole of France, which is not strictly 

 true, but the error resulting from this mode of treatment 

 is probably not greater than the errors of the observa- 

 tions themselves. 



The final values are then tabulated and compared 

 with the values obtained for the same places as deduced 

 from the curves given by Lamont, and in this way a 

 measure of the secular change is obtained. The results 

 are finally plotted in the form of maps on Mercator's 

 projection, giving lines of equal declination, force (hori- 

 zontal component), and dip, and there is, lastly, a map of 

 magnetic meridians. As to the methods employed in 

 the construction of these maps there are unfortunately no 



details. It would seem that the lines are simply free- 

 hand curves, so drawn as to best represent the observa- 

 tional results. There is at least no evidence that the 

 results have been combined, as is the practice among 

 English magneticians, so as to obtain the most probable 

 direction of the lines by calculation, and therefore inde- 

 pendently of bias on the part of the map-maker. M. 

 Moureaux moreover offers us no direct means of com- 

 paring the values as taken out from his curves with the 

 actual values obtained at the various stations.- The 

 maps, however, show certain points of interest which 

 may be thus briefly summarized : — 



(i) In the north of France the declination varies about 

 30' for each degree of longitude ; this proportion decreases 

 in the south. The difference in declination between two 

 points at a given distance apart on the same parallel 

 increases with the latitude, and the isogonal lines are 

 closer together in the north than in the south. The most 

 remarkable feature in the declination map is the form of 

 the curves in Brittany and more especially in the neigh- 

 bourhood of Rennes. Their regularity is broken in such 

 manner as to suggest that they are modified by the par- 

 ticular trend of the coast-line. Throughout the whole of 

 the north-west portion of France the declination is less 

 than would be expected from the direction and character 

 of the lines over the rest of the Continent. A comparison 

 with Lamont's map for 1854 shows that the declination 

 has diminished during the thirty years by about 3" 58' in 

 the north, and by about 3° 19' in the south of France. 

 The mean annual decrease in declination seems to 

 increase pretty regularly from south-south-east to north- 

 north-west, or in a direction approximating to that of the 

 magnetic north ; hence the curves of equal declination 

 have not been displaced, by time, parallel to themselves, 

 but have gradually approached to the direction of the 

 geographical meridian. 



(2) The map of lines of equal horizontal component 

 shows that the minimum, 018460 (C.G.S. units), is 

 observed at Dunkirk, and the maximum, o'22i24, at 

 Perpignan, or a difference of 003654 for the interval of 

 8^ of latitude which separates the two points. The 

 maximum rate of decrease of the horizontal component 

 takes place in a direction approximating to that of the 

 magnetic meridian. The decrease is more rapid in the 

 south than in the north, and the interval between two 

 consecutive curves increases pretty regularly with the 

 latitude. The direction of these lines, like those of 

 declination, seems to be modified towards the north-west 

 of France, in such manner that the line corresponding to 

 o'i9o is nearly straight and does not bend to the south as 

 do the others. At places in the extreme north-west of 

 France the value of the horizontal component is therefore 

 greater than the general direction of the other lines 

 would indicate should be the case. A comparison with 

 Lamont's map for 1848 shows that the horizontal com- 

 ponent has increased from about oooS to 0010 in abso- 

 lute value during the thirty-six years. The lines of equal 

 horizontal component have not been displaced parallel to 

 themselves, but are more inclined towards the east, so as 

 to approach the direction of the geographical parallels. 

 The secular change is at its maximum in the west, and 

 diminishes slightly towards the east. 



(3) The map of isoclinals shows that these lines have 

 sensibly the same orientation as the lines of equal hori- 

 zontal component ; i.e. these are very nearly normal to 

 the direction of the magnetic needle. Whilst the inclina- 

 tion diminishes in general towards the south, the interval 

 between two consecutive curves decreases pretty regu- 

 larly with the latitude. The direction of the lines 

 corresponding to 66° and 67' seems to be slightly 

 modified as they cross the north-west part of France, as 

 are the lines of equal horizontal component. During the 

 thirty-six years which have intervened since the date of 

 Lamont's map, the dip has decreased by about i'' 35' in 



